M05 - Part 1_ Discussion - Network Architecturemo
.docx
keyboard_arrow_up
School
Ball State University *
*We aren’t endorsed by this school
Course
101
Subject
Computer Science
Date
Feb 20, 2024
Type
docx
Pages
16
Uploaded by MagistrateHawkMaster3759
Collapse Subdiscussion
Noah Caslin
Noah Caslin
Monday
Feb 12 at 12:30pm
Manage Discussion Entry
In spine-and-leaf architecture: the oversubscription ratio describes the relationship between the bandwidth of leaf switches connecting to servers and the bandwidth of those switches' uplinks to spine switches. An appropriate oversubscription ratio, typically 3:1 or lower, balances cost with performance since some level of oversubscription is acceptable based on average traffic patterns. This ratio helps determine the number of leaf switches supportable in a network considering their uplink capacity. The number of spine switches needed depends on factors such as their port density (higher density allows for more connected leaf switches), desired redundancy levels (e.g., N+1), overall traffic patterns, and the use of technologies like ECMP to balance traffic distribution. Spine-and-
leaf design is flexible, and network traffic should be monitored for potential adjustments over time.
Reply
Reply to Comment
Collapse Subdiscussion
Karen Sears
Karen Sears
Tuesday
Feb 13 at 9:43pm
Manage Discussion Entry
Noah, your explanation provides a clear and concise overview of the oversubscription ratio in spine-and-leaf architecture, as well as considerations for determining the number of leaf and spine switches in a network. It serves as an informative resource for those looking to grasp the essentials of designing and managing networks with this architecture. Well done! Karen Reply
Reply to Comment
Collapse Subdiscussion
Robert D Lasalle
Robert D Lasalle
Thursday
Feb 15 at 9:19pm
Manage Discussion Entry
Noah.
You did a really good job explaining Spine and Leaf Architecture. I agree with Karen about adding depth to the explanation. I like the spine and leaf design due to it flexibility when adding or removing from it. Your explanation is easy to understand. outstanding!! Reply
Reply to Comment
Collapse Subdiscussion
Corinna Valenti
Corinna Valenti
Yesterday
Feb 17 at 5:33pm
Manage Discussion Entry
Noah, I agree with the others that your explanation of the relationships between spine and leaf and oversubcription ratio was done well. I think
it was a great idea to add in the different factor that come into play with
the number of spine switches needed. I definitely had to reread this module twice. It was not an easy one for me. Thank you for breaking it down further. Reply
Reply to Comment
Collapse Subdiscussion
Dalton James Shad May
Dalton James Shad May
5:20pm
Feb 18 at 5:20pm
Manage Discussion Entry
Hello Noah
I had similar conclusions in my own research. The main difference is I found that it is a 2;1 ratio. However this still works with your findings.
Reply
Reply to Comment
Collapse Subdiscussion
Jose Fuentes
Jose Fuentes
Tuesday
Feb 13 at 9:35pm
Manage Discussion Entry
The oversubscription ratio ensures that there is no excessive bandwidth contention when all servers send traffic simultaneously. An appropriate oversubscription ratio is 3:1 or less, and this helps you choose the number of leaf switches in a network because port density on the spine limits the maximum number of leaf switches. So, in return port density and leaf throughput determines how many spine switches you have.
Reply
Reply to Comment
Collapse Subdiscussion
Karen Sears
Karen Sears
Tuesday
Feb 13 at 9:46pm
Manage Discussion Entry
Jose, You mention of port density and leaf throughput influencing the number of spine switches adds depth to the explanation. This reflects the practical considerations that network designers must account for, ensuring that the spine switches can effectively handle the traffic from the leaf switches. The emphasis on industry-appropriate oversubscription ratios and the influence of port density on spine switches enhances the understanding of key considerations in network
design. Well done! Karen Reply
Reply to Comment
Collapse Subdiscussion
Caleb Lee Hendey
Caleb Lee Hendey
Wednesday
Feb 14 at 1:02pm
Manage Discussion Entry
Jose, It's clear that keeping the oversubscription ratio at 3:1 or lower helps prevent too much traffic congestion. Your point about how the number of leaf switches depends on the port density of the spine switches makes sense. Understanding this connection is key for making sure the network runs smoothly. Keep up the great insights!
Edited by Caleb Lee Hendey on Feb 14 at 1:02pm
Reply
Reply to Comment
Collapse Subdiscussion
Karen Sears
Karen Sears
Tuesday
Feb 13 at 9:41pm
Manage Discussion Entry
The oversubscription ratio refers to the relationship between the total bandwidth provided by the network's uplinks (typically from leaf switches) to the core (spine switches) and the aggregate bandwidth
of the devices connected to the network.
An oversubscription ratio of 1:1 implies that the aggregate bandwidth of the access switches equals the total bandwidth of the uplinks. However, a common and cost-effective oversubscription ratio is often in the range of 10:1.
This means the total bandwidth of the uplinks is ten times greater than the aggregate bandwidth of the connected devices.
To determine the number of leaf switches needed, consider the total bandwidth requirements of the connected devices, and choose leaf switches that, in combination, can handle that load while adhering to the chosen oversubscription ratio. This approach ensures efficient utilization of resources while maintaining acceptable performance levels.
Determining Spine Switches: The number of spine switches needed in a spine-and-leaf network is influenced by factors such as the total number of leaf switches, the desired oversubscription ratio, and the
overall performance and redundancy requirements.
The number of spine switches (S) can be determined using the formula: S = sqrt(N), where N is the total number of leaf switches.
If there are 16 leaf switches, the number of spine switches required would be the square root of 16, which is 4. Therefore, a network with 16 leaf switches would typically have 4 spine switches in a full-mesh topology.
Reply
Reply to Comment
Collapse Subdiscussion
Michaley Kinser
Michaley Kinser
Yesterday
Feb 17 at 1pm
Manage Discussion Entry
Karen,
I like your explanation of determining spine switches and leaf switches and how you provided the formula to do so. Who knew there were so many factors that went into networking!? I sure didn't before starting this class! Reply
Reply to Comment
Collapse Subdiscussion
Corinna Valenti
Corinna Valenti
Yesterday
Feb 17 at 5:35pm
Manage Discussion Entry
Hello Karen, You did an amazing job of breaking down how to determine the number of spine and leaf switches needed. The article I went off had a 3:1 ratio, although I did see some with the 10:1 ratio like yours. Thank you for your explanation. You did a great job!
Reply
Reply to Comment
Collapse Subdiscussion
Francisco Ruiz
Francisco Ruiz
Yesterday
Feb 17 at 8:07pm
Manage Discussion Entry
Your explanation of the oversubscription ratio and its impact on network design is clear and concise. The oversubscription ratio indeed plays a crucial role in determining the balance between available bandwidth and the total bandwidth demanded by connected devices.
Your mention of the common oversubscription ratio being around 10:1 aligns with industry standards and reflects the trade-off between cost-
effectiveness and network performance. By explaining how this ratio is calculated and its significance, you've provided valuable insight into network planning.
Additionally, your approach to determining the number of leaf switches by considering the total bandwidth requirements of connected devices is practical and effective. It ensures that network resources are utilized efficiently while maintaining satisfactory performance levels, which is essential for overall network optimization.
Moreover, your explanation of determining the number of spine switches using the square root of the total number of leaf switches is straightforward and easy to understand. This formula provides a quick and efficient method for scaling spine-and-leaf architectures based on the size of the network.
Overall, your response demonstrates a solid understanding of spine-
and-leaf architecture, oversubscription ratio, and the factors influencing
the determination of leaf and spine switches. Keep up the excellent work! Reply
Reply to Comment
Collapse Subdiscussion
Robert D Lasalle
Robert D Lasalle
Tuesday
Feb 13 at 9:55pm
Manage Discussion Entry
Hello Class,
Anyone feeling like our tiny little brains are absorbing a massive amount if information in a VERY short amount of time? Here is the answers to Spine and Leaf Architecture. I found there was a ton of really good information and great examples doing research!
What is oversubscription ratio?
“This is defined as the maximum amount of traffic a switch can receive from end points (i.e., servers directly attached to the switch), divided by the maximum amount of traffic it can send on the uplinks to the rest of the data centre network.” (Lyons & Zolotow, 2020)
What is an appropriate oversubscription ratio for a network?
Doing research there were a lot of different answers. I find that an appropriate oversubscription ratio could be 3:1 or even less 2.5:1. This is a very good example to calculate the ratio. I thought this was worth sharing:
“The oversubscription ratio can be calculated using the formula: (Pn x Ps):
(Un x Us) Pn is the number of connected leaf ports and Ps is their speed of these ports and Un is the number of spine uplinks and Us is their speed. As an example, imagine a Leaf switch with 48x10g ports for attaching endpoints giving 480Gb/s of port capacity. If this Leaf is connected to 4 spine switches at 40Gb/s it will have a total uplink capacity of 160Gb/s. So the oversubscription ratio is: 480:160 and dividing both sides by 160 we get 3:1.” (Lyons & Zolotow, 2020)
How does this help you choose the number of leaf switches in a network? The number of leaf switches is determined by how many ports are available on the spine switch.
How do you determine how many spine switches a network needs? The number of spine switches is governed by the combination of the required throughput between leaf switches, the number of redundant paths, and their port density.
Lyons, M., & Zolotow, C. (2020, September 7). Converged Networking: Over-
subscription ratio
. Medium. https://mikenovahunter.medium.com/converged-
networking-over-subscription-ratio-f1c74348a5f3 Reply
Reply to Comment
Collapse Subdiscussion
Caleb Lee Hendey
Caleb Lee Hendey
Wednesday
Feb 14 at 1:06pm
Manage Discussion Entry
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
- Access to all documents
- Unlimited textbook solutions
- 24/7 expert homework help
Related Questions
CRC (Cyclic Redundancy Check) is done at this layer of the OSI model:
Session
Physical
Application
Data Link
arrow_forward
(c) Figure 1 shows the routing in a wireless sensor network from source sensor node 0 to the
destination sensor node 12.
12
8
7
11
10
3
5
4
2
Figure 1: A wireless sensor network
Assume either DSR or DSDV routing protocol is used in this sensor network. The link between
sensor nodes 3 and 11 suddenly is not functional due to a faulty device. Analyse how the routing
protocol recovers from link breakage and finds a new route.
arrow_forward
please solve this, thanks.
arrow_forward
assume that T = a
arrow_forward
Virtual circuit networks use the phrase "connection status information" to describe the state of a connection. A switch's forwarding database must be updated at what pace it creates and removes connections in a VC network at a rate of one connection per millisecond (on average).
arrow_forward
On Ethernet networks, data is delivered to devices by their MAC addresses. For this to happen, switches and PCs dynamically build ARP caches and MAC address tables. With only a few computers on the network this process seems fairly easy. What might be some of the challenges on larger networks?
arrow_forward
An Internet-service provider company uses a fiber-optic network to transmit data between locations. Data are carried through fiber-optic cables and switching nodes. A portion of the company's transmission network is shown here. The numbers above each arc
show the capacity in gigabits of data per second that can be transmitted over that branch of the network.
3
4
3
4
2
4
1
2
6
5
5
3
2
7
To keep up with the volume of information transmitted between origin and destination points, use the network to determine the maximum amount of data, in gigabits per second, that may be sent from node 1 to node 7.
X GB/s
1
arrow_forward
The term "connection status information" is used in virtual circuit networks to characterise the state of a connection. The forwarding database of a switch must be updated at a rate of one connection per millisecond when it establishes and removes connections in a VC network (on average).
arrow_forward
Explain how bridges operate at the data link layer of the OSI model.
arrow_forward
Explain the role of switches and bridges at the Data Link layer of the OSI model.
arrow_forward
Computer Science
Two flows are transmitting at a constant rate of 4Mbps and 8Mbps and compete for service at a 10Mbps link. If the link adopts an RED policy, what is the expected drop probability for each flow?
arrow_forward
14. In ethernet ring protection the protected link is selected by…
A: lowest cost setting
B: highest priority setting
C: explicitly defining protection link
D: lowest link speed
15. In STP the following parameters are used directly to determine whether the port will end up as forwarding or blocked:
A: port number
B: bridge MAC address
C: cost setting
D: port priority setting
E: link speed
F: A and B
G: B and D
H: B, C and D
○ I: A, C and D
J: none of the above
arrow_forward
I)“When the store-and-forward switching method is used in a switch, it makes a fast forwarding decision based on the source MAC address of a received frame.” Appraise this statement.
arrow_forward
a)
Based on Figure 2-1, answer the followings:
yain Server
Refeption
Switch R
Bridge ID:
Priority- 28673
Mac Address - 00:23:59:A7:41:10
Fa0/11
Fa0/12
Swch R
Fa0/12
Switch S
Fa0/21
Switch S
Bridge ID:
Priority= 32769
Mac Address - 00:23:59:A7:41:18
Fa0/11,
Fa0/22
Switch T
Bridge ID:
Priority= 32769
Mac Address 00:23:59:A7:41:1A
Switch T
Figure 2-1 A multi-switched network that implements STP
The Spanning Tree Protocol (STP) is a network protocol that builds a loop-free logical topology
for an Ethernet network. Based on Figure 2-1, illustrate and explain the 4-step STP convergence
that would happen as below:
(i)
Root Bridge Election.
(ii)
Determine Root Port(s) on ALL non-root bridges.
(iii)
Determine Designated Port(s) (Only consider segments between non-root bridges).
(iv)
Determine Blocked Port.
arrow_forward
TRUE OF FALSE
Every network device implements the transport layer.
[ ] In the Ring topology, each computer acts as repeater make error detection and boosts the signal before sending it on.
[ ] In bus topology network media is inexpensive and easy to work with, so it is easier to trouble shoot than other topology.
[ ] Error control only detects errors
arrow_forward
A data link layer switch SW1 is connected with four hosts (H1, H2, H3, H4) as shown in figure below:
Suppose switch table is empty and switch receives a frame with source MAC address of H1 (MAC-H1) and destination MAC address of H2 (MAC-H2). What actions will be performed by the switch in this case?
arrow_forward
How does a switch operate at the data link layer of the OSI model?
arrow_forward
Unlike datagram and virtual-circuit networks, circuit-switched networks do not need a routing or switching table to establish the output port from which destination-related data may be sent. Explain the cause of this disparity.
arrow_forward
Explain which OSI model layer switches operate at, and provide an example of a device that operates at this layer.
arrow_forward
course: Data communication and communication Networks
Q:
Compare the different TCP variants (Tahoe, Reno, New Reno, and SACK) forcongestion control and congestion avoidance mechanisms highlighting advantagesand disadvantages. You are also required to discuss the fairness of their congestioncontrol mechanisms using Raj Jain’s Fairness Index.
arrow_forward
Explain how a Layer 2 switch operates at the data link layer in the OSI model.
arrow_forward
Explain the role of a switch in the OSI model, and which layer(s) does it primarily operate at?
arrow_forward
Explain the role of a router in the OSI model, specifying which layer it operates at and its primary function.
arrow_forward
A data link layer switch SW1 is connected with four hosts (H1,
H2, H3, H4) as shown in figure below:
a) Suppose switch table is empty and switch receives a
frame with source MAC address of H1 (MAC-H1) and
destination MAC address of H2 (MAC-H2). What actions
will be performed by the switch in this case?
b) Consider switch MAC address learning capability, how
many entries are expected in switch table
c) Show all entries of switch table
d) Consider the part "a" again, but with a fully populated
switch table (all entries) instead of empty table
e) H1 is unplugged from this LAN by the system
administrator. Does this change affect the switch table?
Explain your answer.
Host 1
Link 1
Link 4
Data-link
Host 4
Host 2
Layer
Link 2
Link 3
Host 3
arrow_forward
SEE MORE QUESTIONS
Recommended textbooks for you
Operations Research : Applications and Algorithms
Computer Science
ISBN:9780534380588
Author:Wayne L. Winston
Publisher:Brooks Cole
Systems Architecture
Computer Science
ISBN:9781305080195
Author:Stephen D. Burd
Publisher:Cengage Learning
Related Questions
- CRC (Cyclic Redundancy Check) is done at this layer of the OSI model: Session Physical Application Data Linkarrow_forward(c) Figure 1 shows the routing in a wireless sensor network from source sensor node 0 to the destination sensor node 12. 12 8 7 11 10 3 5 4 2 Figure 1: A wireless sensor network Assume either DSR or DSDV routing protocol is used in this sensor network. The link between sensor nodes 3 and 11 suddenly is not functional due to a faulty device. Analyse how the routing protocol recovers from link breakage and finds a new route.arrow_forwardplease solve this, thanks.arrow_forward
- assume that T = aarrow_forwardVirtual circuit networks use the phrase "connection status information" to describe the state of a connection. A switch's forwarding database must be updated at what pace it creates and removes connections in a VC network at a rate of one connection per millisecond (on average).arrow_forwardOn Ethernet networks, data is delivered to devices by their MAC addresses. For this to happen, switches and PCs dynamically build ARP caches and MAC address tables. With only a few computers on the network this process seems fairly easy. What might be some of the challenges on larger networks?arrow_forward
- An Internet-service provider company uses a fiber-optic network to transmit data between locations. Data are carried through fiber-optic cables and switching nodes. A portion of the company's transmission network is shown here. The numbers above each arc show the capacity in gigabits of data per second that can be transmitted over that branch of the network. 3 4 3 4 2 4 1 2 6 5 5 3 2 7 To keep up with the volume of information transmitted between origin and destination points, use the network to determine the maximum amount of data, in gigabits per second, that may be sent from node 1 to node 7. X GB/s 1arrow_forwardThe term "connection status information" is used in virtual circuit networks to characterise the state of a connection. The forwarding database of a switch must be updated at a rate of one connection per millisecond when it establishes and removes connections in a VC network (on average).arrow_forwardExplain how bridges operate at the data link layer of the OSI model.arrow_forward
- Explain the role of switches and bridges at the Data Link layer of the OSI model.arrow_forwardComputer Science Two flows are transmitting at a constant rate of 4Mbps and 8Mbps and compete for service at a 10Mbps link. If the link adopts an RED policy, what is the expected drop probability for each flow?arrow_forward14. In ethernet ring protection the protected link is selected by… A: lowest cost setting B: highest priority setting C: explicitly defining protection link D: lowest link speed 15. In STP the following parameters are used directly to determine whether the port will end up as forwarding or blocked: A: port number B: bridge MAC address C: cost setting D: port priority setting E: link speed F: A and B G: B and D H: B, C and D ○ I: A, C and D J: none of the abovearrow_forward
arrow_back_ios
SEE MORE QUESTIONS
arrow_forward_ios
Recommended textbooks for you
- Operations Research : Applications and AlgorithmsComputer ScienceISBN:9780534380588Author:Wayne L. WinstonPublisher:Brooks ColeSystems ArchitectureComputer ScienceISBN:9781305080195Author:Stephen D. BurdPublisher:Cengage Learning
Operations Research : Applications and Algorithms
Computer Science
ISBN:9780534380588
Author:Wayne L. Winston
Publisher:Brooks Cole
Systems Architecture
Computer Science
ISBN:9781305080195
Author:Stephen D. Burd
Publisher:Cengage Learning